import os
import torch
from PIL import Image
import torchvision.transforms.functional as F
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.patches as patches

# Visualization function for predictions (bounding boxes only)
def visualize_predictions(image_path, model, threshold=0.05):
    model.eval()
    img = Image.open(image_path).convert('RGB')
    img_tensor = F.to_tensor(img).unsqueeze(0).to(device)

    # Perform inference
    with torch.no_grad():
        outputs = model(img_tensor)

    predictions = outputs[0]
    pred_boxes = predictions['boxes'].detach().cpu().numpy()
    pred_scores = predictions['scores'].detach().cpu().numpy()

    # Filter predictions based on confidence threshold
    keep = pred_scores >= threshold
    pred_boxes = pred_boxes[keep]

    if len(pred_boxes) == 0:
        print(f"No predictions above the threshold for {image_path}!")
        return

    # Visualize the results
    fig, ax = plt.subplots(1, figsize=(12, 8))
    ax.imshow(img)
    for box in pred_boxes:
        xmin, ymin, xmax, ymax = box
        rect = patches.Rectangle(
            (xmin, ymin), xmax - xmin, ymax - ymin,
            linewidth=2, edgecolor='red', facecolor='none'
        )
        ax.add_patch(rect)

    plt.axis('off')
    plt.show()  # Display the image with predictions
    print(f"Displayed visualization for {image_path}")

# Recursive function to process a directory (bounding boxes only)
def process_folder(input_folder, model, threshold=0.05):
    for root, _, files in os.walk(input_folder):
        for file in files:
            if file.lower().endswith(('.jpg', '.jpeg', '.png')):
                image_path = os.path.join(root, file)
                print(f"Processing {image_path}")
                visualize_predictions(image_path, model, threshold)

# Paths to model and data
model_path = "mask_rcnn_epoch_100.pth"  # Path to your saved model weights
input_folder = "/home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images"  # Path to input folder
num_classes = 2  # Number of classes including background

# Load the model
model = load_trained_model(model_path, num_classes)

# Process the folder recursively and display images
process_folder(input_folder, model, threshold=0.0695)

/tmp/ipykernel_95777/3553455240.py:18: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  model.load_state_dict(torch.load(model_path, map_location=device))

Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_10_0.jpg

Displayed visualization for /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_10_0.jpg
Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_5_4.jpg

Displayed visualization for /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_5_4.jpg
Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_3_2.jpg

Displayed visualization for /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_3_2.jpg
Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_2_3.jpg

Displayed visualization for /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_2_3.jpg
Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_4_2.jpg
No predictions above the threshold for /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_4_2.jpg!
Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_20_2.jpg

Displayed visualization for /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_20_2.jpg
Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_18_4.jpg

Displayed visualization for /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_18_4.jpg
Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_3_4.jpg

Displayed visualization for /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_3_4.jpg
Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_4_1.jpg
No predictions above the threshold for /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_4_1.jpg!
Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_29_4.jpg

Displayed visualization for /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_29_4.jpg
Processing /home/brt_ai/PROJECTS/maskrcnn/datset_full_raw/val/images/aug_9_0.jpg